Singapore’s Digital Urban Climate Twin

The Digital Urban Climate Twin (DUCT) is the core outcome of Cooling Singapore 2.0, developed to help planners understand how different policies and interventions affect the urban climate. By integrating models of urban climate, energy systems, and anthropogenic heat, DUCT allows users to simulate “what-if” scenarios and explore strategies to mitigate the Urban Heat Island (UHI) effect, improve outdoor thermal comfort, and enhance energy efficiency.

Singapore also operates Virtual Singapore, a 3D digital model/digital twin platform integrating geospatial, building, and infrastructure data (since 2014, completed by 2022), which serves as an enabling infrastructure for multiple applications including DUCT.

Beyond Cooling Singapore 2.0, the work on the Digital Urban Climate Twin continues under Future Cities Laboratory Phase IV (2026–2030). This next phase aims to reimagine cities as settlement systems and engines of positive change, shaping a sustainable and equitable future across four priority domains: health, biodiversity, community, and decarbonisation. The approach connects urban climate modelling with wider systemic challenges such as habitat and biodiversity loss, droughts and floods, food shortages, extreme weather, political instability, vector-borne diseases, social inequalities, population growth, increased resource demand, waste, transport, mining and imports, and climate-related risks. The DUCT contributes as a foundational digital tool supporting this broader integrated systems view of cities.

Implementing Singapore’s Digital Urban Climate Twin required substantial technical, institutional, and design work. Below is a breakdown of the approach:

A. Model Integration & Architecture

• Combine meso-scale climate modeling (e.g. WRF) and micro-climate models with urban models (building energy, traffic, land-use/ vegetation) and anthropogenic heat emission models (buildings, transport, industry).
• Structure as a “Simulation-as-a-Service” federation: modular models, cloud or HPC execution, and standardized interfaces.
• Develop the DUCT Explorer, a browser-based GUI that abstracts modelling complexity and allows users to set scenarios (e.g. increase green cover, change building efficiency, EV uptake) without coding.

B. Data Input & Calibration

• Ingest multi-source data: GIS, remote sensing, building typologies, vegetation, traffic flows, energy usage, meteorological data.
• Calibrate and validate model outputs against observed data (weather stations, microclimate sensors), to ensure reliability.
• Iteratively refine parameters and modular coupling (e.g. heat coupling among subsystems) to ensure stable, credible simulation results.

C. Use-case Development & Scenario Design

• Co-develop scenario templates aligned with Singapore’s Green Plan 2030 (e.g. EV share 18%, increased vegetation, better building efficiency) for modeling demonstration.
• Tailor scenarios at different scales: island-wide, district-level, neighborhood-level, or even street-level interventions.

D. Stakeholder Engagement and Pilot Testing

• Conduct guided workshops and sessions with urban planners and government agencies (Q4 2023, Q1 2024) to test usability, collect feedback, and refine user interface and workflows.
• Adapt the DUCT Explorer and scenario library based on user needs, ensuring relevance to planning decision workflows.

E. Performance & Readiness Upgrades

• Focus on robustness, computational efficiency, and scalability to support full city-wide runs.
• Aim for Technology Readiness Level (TRL) approx. 6+ by mid-2024.
• Expand flexibility to onboard additional modules/domains (e.g. energy systems, climate adaptation).

The success of DUCT depends on multi-institutional collaboration:

• Singapore-ETH Centre leads the project; key collaborators include Singapore Management University (SMU), Singapore-MIT Alliance for Research and Technology (SMART), TUMCREATE (established by the Technical University of Munich), National University of Singapore (NUS), and Cambridge CARES.
• Funded via Singapore National Research Foundation, Prime Minister’s Office, under the CREATE programme.
• Government agencies (e.g. Ministry of National Development, Urban Redevelopment Authority, NParks) contribute domain insight, data, planning alignment.
• Academic & research institutions supply modeling, climate, data science, urban planning expertise
• Engagement with planners via workshops, collaborative design of scenario use cases

Effective communication and awareness are key to adoption:

• To build understanding and trust in DUCT’s value, the project uses multiple channels to communicate results, methods, and use cases.
• Develop visualizations and dashboards (via the Explorer) that can be shared with non-technical stakeholders (e.g. politicians, public)
• Present case study results (e.g. cooling maps) in public forums, conferences, and government briefings
• Co-publish reports, policy briefs, academic articles to disseminate learnings
• Engage planners in workshops demonstrating how DUCT can support their decisions
• Use simplified scenario stories (e.g. before/after maps) to illustrate benefits

Financing and anchoring resources are critical:

• Core funding from National Research Foundation, Singapore (CREATE programme) for Cooling Singapore phases.
• In-kind contributions: computational infrastructure, data from government agencies, institutional support from partner universities
• Human capital: multidisciplinary researchers, modelers, UX designers, computational specialists
• Aim to scale to more operational deployment would require further funding (e.g. from governmental planning budgets)
• Resources are allocated for model development, system maintenance, stakeholder engagement, capacity building

The project has produced demonstrable modeling insights, informed planning, and proof-of-concept uptake as it transitions from research to operational use.

• The modeling produced quantified estimates of how cooling strategies (greening, EVs, building improvements) might affect urban temperatures (e.g. up to 0.13 °C cooling island-wide)
• Insights revealed that buildings (especially passive thermal mass and AC exhaust) have significant contributions to urban heat, enabling targeted interventions.
• Planners now can test alternative policy or design options (e.g. vegetation placement, EV uptake, building retrofits) before implementation
• The user interface (DUCT Explorer) lowers the barrier for non-technical users to engage with complex models
• Enhanced capacity in government agencies to integrate climate-aware urban planning
• Although still in pilot/experimental stage, the project has helped Singapore embed heat-resilience thinking into Green Plan 2030 policy dialogues.

The project also faced significant challenges, which provide lessons for others.

• Data gaps and quality: Not all data sources have sufficient spatial/temporal resolution or completeness; integration across domains is non-trivial.
• Model coupling complexity: Ensuring stability, interoperability, and consistency between climate, building, traffic, and heat-emission models is technically challenging.
• Computational cost & scalability: Running high-resolution city-wide simulations demands high-performance computing and optimization.
• Usability for planners: Bridging the gap between model complexity and usability for non-experts requires careful interface design and iteration.
• Uncertainty and sensitivity: Projections depend on assumptions, input uncertainties, and scenario definitions; interpreting results for policy must consider uncertainty bounds.
• Adoption inertia: Convincing agencies to shift planning practices and trust model outputs takes time and institutional buy-in.
• Funding for operationalization: Sustaining the platform beyond research phase into long-term operational and maintenance funding is a challenge.

Singapore’s DUCT offers a clear model for other cities aiming to integrate urban climate modelling into planning:

• Modular architecture allows adaptation: cities can onboard comparable modules (climate, energy, traffic) relevant to their context.
• Scenario templating aligned with national policy (e.g. climate targets) helps link modeling to decision-making.
• User-friendly interfaces are critical: non-modelers must be able to run and interpret scenarios.
• Strong data ecosystems are required: cities need geospatial, energy, traffic, meteorological, land-use, vegetation data.
• Institutional anchoring and partnerships: bridging academia, government, planning agencies is essential for legitimacy, data access, and uptake.
• Phased deployment: start with pilot districts, validate, build trust, then scale to city-wide.
• Sustainability and funding plan: plan for ongoing maintenance, updates, and capacity building beyond initial development.

In conclusion, Singapore’s Digital Urban Climate Twin demonstrates how integrated modeling, thoughtful interface design, and institutional collaboration can transform how cities plan for heat resilience and climate adaptation. With careful adaptation to local contexts (climate, data availability, institutional capacity), this approach is replicable in many cities globally facing heat stress challenges.

• “A Digital Urban Climate Twin for Singapore” (SSRN) (papers.ssrn.com)
• Cooling Singapore / Singapore-ETH Centre project page (Singapore-ETH Centre)
• Singapore’s Digital Urban Climate Twin (World Cities Summit poster) (worldcitiessummit.com.sg)
• Cooling Singapore 2.0 details (NUS) (fass.nus.edu.sg)
• URA “Shaping a Heat Resilient City” overview (Urban Redevelopment Authority)

Ander Zozaya
Singapore-ETH Centre
Cooling Singapore